CN105116307B - A kind of screening method of gas insulated combined electrical equipment status early warning data outliers - Google Patents

Abstract

The invention belongs to the research and application technical field of partial discharge state monitoring and fault diagnosis of gas-insulated combined electrical appliances in smart grids, and particularly relates to a screening method for abnormal value of state early warning data of gas-insulated combined electrical appliances. The monitoring device collects the monitoring data of each set of gas-insulated combined electrical appliances, and uploads the monitoring data to the data server of the main station system platform of the monitoring and early warning center for data analysis; Abnormal values in the partial discharge early warning data of electrical state characteristics; and accurately identify the typical abnormal value representation in the discharge amplitude, avoiding misjudgment of monitoring early warning and fault diagnosis caused by abnormal values, and finally reliably evaluating the state of the equipment, solving the comprehensive Grasp and distinguish the monitoring data of equipment and its environment in real time; it is helpful to deal with the faults of combined electrical equipment widely distributed in various places.

Description

A Screening Method for Abnormal Values of State Early Warning Data of Gas Insulated Combined Electrical Appliances

technical field

The invention belongs to the technical field of research and application of partial discharge state monitoring and fault diagnosis of gas-insulated combined electrical appliances in smart grids, and in particular relates to a method for screening abnormal values of state early warning data of gas-insulated combined electrical appliances.

Background technique

With the progress of online monitoring technology at home and abroad, more and more online monitoring systems for substation equipment have been put into operation, which can play a positive role in timely grasping the operating status of key substation equipment and improving the safe operation level of equipment. Among them, the on-line partial discharge monitoring technology of gas-insulated combined electrical appliances, which is a relatively mature technology and widely used monitoring technology; and the application of this technology has achieved many successful cases in the power industry. In the power industry, the on-line partial discharge monitoring technology of gas-insulated combined electrical appliances mainly uses the live/on-line detection method based on UHF technology to identify the discharge signal inside the equipment, diagnose the degree of insulation deterioration inside the equipment, and further judge whether there is a basin insulator in the equipment body Faults such as breakdown or sulfur hexafluoride gas breakdown. First of all, online monitoring devices distributed throughout the province are continuously collecting and detecting characteristic parameters (monitoring data) such as the amplitude of partial discharge of gas-insulated composite electrical appliances; and sending the monitoring data to the monitoring and early warning center every 15 minutes. Secondly, through the monitoring and early warning center of the provincial power grid, the monitoring data such as the discharge amplitude of more than hundreds of partial discharge monitoring devices are collected and processed in parallel. Thirdly, judge the status of equipment based on the changes in partial discharge monitoring data, and formulate equipment early warning and operation and maintenance strategies. In view of this, the technicians in the monitoring and early warning center need to investigate and analyze the abnormal values in the monitoring data in the first time. The monitoring and early warning centers of the provincial power grids are also the first defense system to ensure the safety and stability of gas-insulated combined electrical appliances. The role is extremely important.

The partial discharge monitoring device can detect parameters such as the amplitude, frequency, and phase of partial discharge, and the amplitude of the discharge is an important indicator for judging the health status of the equipment. The discharge amplitude of gas-insulated combined electrical appliances in a normal state is normally distributed (Gaussian distribution), and its change curve is the central value of the monitoring data sample, which gradually changes uniformly (fluctuates) to both sides of the size. However, on the one hand, due to the aging of electronic components, complex electromagnetic wave background noise, sensor failure and other influencing factors, the monitoring data often deviates and even exceeds the warning value; thus causing misjudgment and hindering the due effectiveness of monitoring and warning. On the other hand, the monitoring and early warning center must simultaneously process multiple types of massive monitoring data and generate alarm events, but there is still insufficient basis for judging whether there is deviation in the monitoring data; in addition, the operating level of some operation and maintenance personnel needs to be improved, and the diagnosis and analysis experience is insufficient. , it is also difficult to quickly and accurately distinguish background noise and identify latent discharge phenomenon defects in a short period of time.

Based on the above problems, it is currently not possible to use the internal relationship between these real-time monitoring data to confirm the accuracy of the data, nor can it efficiently filter potential interference data or misdiagnosis data; this also increases the difficulty of the operation and maintenance monitoring and early warning center, affecting In order to make full use of monitoring data to guide the operation and maintenance of gas-insulated combined electrical appliances, it is not conducive to consolidating the level of reliable power supply. The fundamental reason is that a large amount of monitoring data contains abnormal values caused by some measurement deviations, which restricts the quality of monitoring data. Therefore, there is an urgent need to develop an economical and practical consistency check and screening method suitable for real-time processing of abnormal values of partial discharge early warning data in the case of large data, and to realize real-time and effective processing of massive concurrent alarms.

Contents of the invention

The purpose of the present invention is to solve the above-mentioned problems in the prior art, and to provide a screening method for abnormal values in the early warning data of the partial discharge state of gas-insulated combined electrical appliances by performing consistency check on the monitoring data of the status of gas-insulated combined electrical appliances Method, in order to realize above-mentioned object, the technical scheme that the present invention adopts is as follows:

A method for screening abnormal values of state early warning data of gas-insulated combined electrical appliances, characterized in that it includes the following steps:

Step 1: Collect the monitoring data (discharge amplitude) of each gas-insulated combined electrical appliance through partial discharge monitoring devices distributed everywhere;

Step 2: Upload the monitoring data step by step to the data server of the main station system platform of the monitoring and early warning center for data analysis;

Step 3: Screen and analyze abnormal values in various partial discharge early warning data that characterize the state characteristics of gas-insulated combined electrical appliances through consistency inspection;

Step 4: Accurately identify the typical abnormal value representation in the discharge amplitude, avoid the interference and misjudgment of abnormal values on normal monitoring and early warning, fault diagnosis, and finally reliably evaluate the status of the equipment.

Preferably, the steps of data analysis performed by the master station system platform data server of said step 2 are as follows:

(1) Analyze and calculate the monitoring data samples in the data server of the main station system platform, and extract the discharge amplitude monitoring data samples of a set of monitoring devices within a specified time period as X1, X2, X3, ..., Xn;

(2) Arrange the monitoring data samples Xi into a statistical data set X(i) in order from small to large, and arrange them randomly if there are monitoring data with the same measured value, that is: X(1)≤X(2)≤...≤ X(n);

(3) Solve the statistic C for different sample sizes n, combine the Dixon analysis method to test the consistency of the historical samples of the discharge amplitude in engineering applications, and solve the statistical value of the confidence interval to obtain a significant difference level α The empirical value is 0.01 , to solve its critical value C1-α(n), so that the sample data value can warn the operating state of the gas-insulated combination device;

(4) Analyzing and diagnosing the status of the monitored gas-insulated combined electrical appliances based on the screened monitoring data samples.

Preferably, the partial discharge monitoring device sends the monitoring data (discharge amplitude) collected in the gas-insulated combined electrical appliance step by step to the data server of the main station system platform through the wired power integrated data network for screening and analysis.

Preferably, the partial discharge monitoring device includes a UHF signal sensor, a signal collector and a comprehensive processing unit, and the monitoring data output by the UHF signal sensor is monitored and analyzed through the signal collector and the comprehensive processing unit in sequence.

The partial discharge monitoring device refers to the continuous measurement of partial discharge that may occur inside the GIS through the external partial discharge UHF signal sensor when the gas-insulated combined electrical equipment is in operation. Usually, the GIS monitoring system is fixedly installed on site and runs online for a long time without human participation, and automatically completes functions such as detection, storage, communication and diagnosis. Among them, the UHF signal sensor refers to the electromagnetic wave signal generated by the sensor monitoring gas-insulated composite electrical measuring equipment in the UHF frequency band (300MHz～3GHz).

In summary, the present invention collects the on-line monitoring data of various gas-insulated combined electrical appliances distributed in a wide area, and uses the Dixon test method to detect one or more abnormal values in a set of measured values in the monitoring and early warning center. , and then based on the diagnosis and analysis of the prior data of the partial discharge status of the gas-insulated combined electrical appliances, a screening method for the abnormal value of the early warning data of the gas-insulated combined electrical appliances is provided, which realizes the multi- Dimensional and intensified partial discharge disputes, the present invention has the following beneficial effects due to the adoption of the above technical solutions:

First, the present invention uses the range ratio analysis method to test the normal distribution of discharge amplitudes, and then uses the confidence interval to screen abnormal values, which significantly improves the credibility of the monitoring data;

Second, the present invention adopts the characteristic clustering analysis method of abnormal data, overcomes the key problem of background interference abnormal value in the external operating environment, and then obtains simplified and strict monitoring and early warning results;

Third, the present invention takes the lead in realizing the processing mechanism for massive concurrent monitoring data and its alarm events, and solves the problem of comprehensively and real-time grasping and distinguishing of monitoring data of equipment and its environment; it helps to deal with fault events widely distributed in various equipment.

Description of drawings

In order to more clearly illustrate the technical solutions in the examples of the present invention or the prior art, the accompanying drawings required in the description of the implementation examples or the prior art will be briefly introduced below. Obviously, the accompanying drawings in the following description are only the present invention. For some examples of the invention, those skilled in the art can also obtain other drawings based on these drawings without paying any inventive step.

Fig. 1 is a flow chart of a method for screening abnormal values of state early warning data of gas-insulated combined electrical appliances according to the present invention.

Fig. 2 is a schematic diagram of the principle of an early warning system of a method for screening abnormal values of early warning data of the status of gas-insulated combined electrical appliances according to the present invention.

In the attached drawings, 1-UHF signal sensor, 2-Signal collector, 3-Comprehensive processing unit, 4-Station switch, 5-Master station system platform router, 6-Master station system platform database server, 7-Web Browse server, 8-operator station, 9-engineer station.

Detailed ways

The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the examples of the present invention. Obviously, the described embodiments are only some of the embodiments of the present invention, not all of them. Based on the embodiments of the invention, all other embodiments obtained by persons of ordinary skill in the art without creative efforts fall within the protection scope of the present invention.

As shown in Figure 1, in the present invention, a screening method for the abnormal value of the state early warning data of gas-insulated combined electrical appliances refers to checking the range error between the maximum value and the minimum value in a group of discharge amplitudes through significant difference analysis, and Whether the data uploaded by the monitoring device satisfies the normal distribution, and then filter outliers in the monitoring data according to the confidence interval, the specific method steps are as follows:

Step 1: Collect the monitoring data (discharge amplitude) of each gas-insulated combined electrical appliance through partial discharge monitoring devices distributed everywhere;

Step 2: Upload the monitoring data step by step to the master station system platform data server of the monitoring and early warning center for data analysis; as shown in Figure 2, the partial discharge monitoring device includes a UHF signal sensor 1, a signal collector 2 and Integrated processing unit 3, after the monitoring data collected by the partial discharge monitoring device is processed by the integrated processing unit 3, it is uploaded to the main station system platform data server 6 through the station switch 4 and the main station system platform router 5; the partial discharge The monitoring device sends the monitoring data (discharge amplitude) collected in the gas-insulated combined electrical appliance step by step to the data server 6 of the main station system platform through the wired power integrated data network for screening and analysis; the UHF signal sensor 1 outputs The monitoring data (discharge amplitude) signal is monitored and analyzed through the signal collector 2 and the comprehensive processing unit 3 in sequence, and the output terminal of the UHF signal sensor 1 is connected to the output terminal of the signal collector 2, and the connection forms two or More than two collection devices are input into the comprehensive processing unit 3 for pre-processing and screening for abnormal values of status warning data.

Step 3: Analyze and calculate the monitoring data samples in the data server of the main station system platform, and analyze the abnormal values in various partial discharge early warning data that characterize the status characteristics of gas-insulated composite electrical appliances through consistency inspection screening, including the following steps:

(1) Extract the discharge amplitude monitoring data samples of a certain partial discharge monitoring device within a specified time period as X1, X2, X3, ..., Xn;

(2) Arrange the monitoring data samples Xi from small to large into a statistical number sequence X(i), if there are monitoring data with the same measured value, they will be randomly arranged, that is: X(1)≤X(2)≤...≤ X(n);

(3) Solve the statistic C for different sample sizes n, combine the Dixon analysis method to check the consistency of the discharge amplitude historical samples in engineering applications, and set the arithmetic mean value of the discharge amplitude monitoring data sample sequence as but:

Therefore, the error of a certain discharge amplitude X _i under the same operating conditions is ΔX, then:

If ΔX is too large, the autocorrelation of the error term may be very serious, which will lead to the inconsistency of the estimated parameters, because the discharge amplitude data series is a non-stationary system; test the level of significant difference, and judge whether the confidence is Satisfy;

(4) According to the statistical value of the confidence interval, the empirical value of the significant difference level α is 0.01. When the sample interval of the discharge amplitude monitoring data is [C ₁ , C ₂ ], the confidence of the interval [C ₁ , C ₂ ] Confidence interval with a level of 1-α, C ₁ is the measured value of the upper confidence limit, and C ₂ is the measured value of the lower confidence limit;

(5) Correspondingly, when the arithmetic mean value of the discharge amplitude monitoring data sample sequence is Solve the significant difference level α empirical value α(n) of the nth sample statistic C, and then solve its critical value In this way, the operating status of the gas-insulated combined electrical appliance can be analyzed through whether the sample data is normal; when the sample statistic C≥ critical value Then the sample x(n) is an outlier; when the sample statistic C<critical value Then the sample x(n) is a normal value;

(6) Analyzing and diagnosing the status of the monitored gas-insulated combined electrical appliances based on the screened monitoring data samples. View the displayed basic statistical spectrum (φ-q-n) and two-dimensional spectrum (φ-q, φ-n, q-n); and based on the partial discharge signal discharge amplitude (q), discharge times (n), visual Diagnose the type and severity of partial discharge in terms of electric quantity (q), luminous flux (Φ), discharge phase (φ) and development trend.

Step 4: Accurately identify the typical abnormal value representation in the discharge amplitude, avoid the interference and misjudgment of abnormal values on normal monitoring and early warning, fault diagnosis, and finally reliably evaluate the status of the equipment

As shown in Figure 2, in a preferred embodiment of the present invention for screening and processing the abnormal value of the status early warning data of a gas-insulated combined electrical appliance, the early warning system for the status early warning of a gas-insulated combined electrical appliance includes a UHF signal sensor 1, a signal collector 2. Integrated processing unit 3, station switch 4, master station system platform router 5, master station system platform data server 6, Web browsing server 7, operator station 8 and engineer station 9, the UHF signal sensor collects and discharges After the signal and output discharge amplitude and other signals are transmitted to the data server 6 of the main station system platform through the signal collector 2, the integrated processing unit 3, the station switch 4, and the main station system platform router 5, the abnormality of the screening status and early warning data is checked. value. The master station system platform data server 6 is also connected to the Web browsing server 7, the operator station 8 and the engineer station 9 respectively, and the UHF signal sensor 1, the signal collector 2, the integrated processing unit 3, and the station switch 4 It will be deployed in the substation; among them, the UHF signal sensor 1 is one or more channels (1 to 16 UHF signal sensors can be selected), and is installed on the shell of the gas-insulated combined electrical appliance body, and the integrated processing unit 3, The station-end switch 4 is installed in the on-line monitoring screen cabinet of the relay protection room; in the present embodiment of the invention, the specific quantity of UHF signal sensor 1, signal collector 2, integrated processing unit 3, and station-side switch 4 is based on the substation The number of devices in the device and the structural features of the device body are determined; and the number of the integrated processing unit 3 and the station-side switch 4 is one set for each station. The main station system platform router 5, the main station system platform data server 6, the Web browsing server 7, the operator station 8, and the engineer station 9 will be deployed in the monitoring and early warning center. After the online monitoring device scans the monitoring data such as the internal discharge amplitude of the gas-insulated combined electrical appliance, it uploads it to the data server of the main station system platform of the monitoring and early warning center to screen for abnormal values, and finally provides consultation for relevant equipment operation and maintenance technicians.

In this embodiment, only the monitoring data of the discharge signal is transmitted in the loop from the UHF signal sensor 1 to the integrated processing unit 3, and the signal collector 2 can receive and process at least 16 UHF signal sensors 1 in sequence The detected discharge signal; the UHF signal sensor 1 is a dual-engine Archimedes planar helical antenna, the frequency band of the detection electromagnetic wave signal is between 300MHz and 3GHz, the standing wave ratio is lower than 2, and all non- Metal parts are tightly wrapped with shielding cloth. In addition, the specific number of UHF signal sensors 1 is determined according to the interval and structural characteristics of the device body; the number of signal collectors 2 depends on how many UHF signal sensors are deployed. Depends on sensor 1.

The signal collector 2 is processed based on an ARM Cortex 32-bit microprocessor chip, and is packaged in a GZPD-AD-UHF series advanced engineering plastic shell. The output terminal of the body discharge signal collector 2 is connected to the input terminal of the integrated processing unit 3 Connect to each other through optical fiber cables, and its signal loop can transmit the detected value; the output end of the integrated processing unit 3 and the input end of the station switch 4 are connected to each other through a twisted pair cable, and its signal loop can transmit the discharge amplitude (q), discharge times (N), apparent electric quantity (Q), luminous flux (Φ), discharge phase (φ), spectrogram, time stamp (acquisition, upload) and control commands, etc.; The integrated data network sends various monitoring data to the main station system platform router 5 of the remote monitoring and early warning center, and the station end switch 4 is an H3C S3600 series switch.

The output end of the master station system platform router 5 and the input end of the master station system platform data server 6 are connected to each other through a twisted pair cable, and its signal loop can transmit various monitoring data, and can issue control commands to the monitoring circuit in the substation etc., the master station switch 7 is a router of the H3C SR8800 series, which can realize the load sharing and mutual backup of multiple uplinks through the distributed cross-device link aggregation technology, and improve the reliability of the complete set of network architecture. It increases the utilization rate of link resources and is suitable for monitoring data communication in wide area distribution. Described main station system platform data server 6 stores the various types of monitoring data uploaded by each substation at the first time, and connects respectively the Web browsing server 7, the operator station 8, the engineer station 9 of the monitoring and early warning center through the twisted pair cable, It can also provide retrieval, query and backup services. The data server 6 of the main station system platform is a server of HP 9000 series, which can screen and process the abnormal value of the early warning data of the status of gas-insulated combined electrical appliances.

The Web browsing server 7 can provide technical personnel at all levels in the provincial power grid with query services related to monitoring data of the entire network according to predetermined access rights. Specifically, technicians at all levels connect to the Web browsing server 7 through the HTTP protocol in the integrated power data network to obtain data within the permitted browsing authority, and then send the data to the web server and obtain specific monitoring data content. The operator station 8 can provide data query and diagnosis services for technicians of the monitoring and early warning center. The main content includes: monitoring time, equipment name, discharge type, discharge spectrum, monitoring report, data comparison, data analysis, data list, comprehensive information, current status, change curve, etc. The operator station 9 can provide the engineers and system maintenance personnel of the monitoring and early warning center with query and diagnosis services of the monitoring data of the whole network, and can issue control instructions to the monitoring systems of each terminal in the integrated electric power data network, including calling Data such as the pulse waveform of the detection signal of the device body or the external environment, upgrade the neural network database in the integrated processing unit 3 .

In conjunction with Fig. 1 and Fig. 2, the specific operation process of the present invention is as follows,

1). Deploy the UHF signal sensor 1, signal collector 2, integrated processing unit 3, and station end switch 4 in the early warning system for the status of gas-insulated combined electrical appliances in the 110kV switchyard area of a 110kV substation, and Connect to the power supply; wherein each UHF signal sensor 1 is tightly installed on the surface of the pot insulator of the gas-insulated composite electrical appliance body;

2). The body signal sensor 1 on the surface of the basin-type insulator installed on the body of the gas-insulated combined electrical appliance detects the discharge signal inside the gas-insulated combined electrical appliance, and is collected by the body signal collector 2 and uploaded to the integrated processing unit 5;

3). The comprehensive processing unit 3 filters, stores and re-uploads the monitoring values uploaded by each signal collector 2 .

4). The comprehensive processing unit 3 transmits the monitoring data of discharge equal amplitude to the switch machine 4 at the station end through a twisted pair cable, and then further provides it to the main station system platform in the monitoring and early warning center through the communication link of the integrated power data network Router 5, master station system platform data server 6;

5). After the data server 6 of the main station system platform screens the abnormal values in the status warning data, it distributes the status warning data to the Web browsing server 7, the operator station 8, and the engineer station 9 for display, storage, comparison, analysis and processing , after processing, the monitoring value and spectrogram representing the degree of discharge are displayed, and provided to the technical personnel of the monitoring and early warning center, the equipment operation and maintenance personnel of the power supply bureau, and external experts to diagnose the insulation status of the 110kV three-phase common box-type gas-insulated combined electrical appliance; Through Web browsing server 7, operator station 8, and engineer station 9, status warning data analysis is realized;

6). Analyze the number of discharges N and the average discharge amplitude U detected by the UHF signal sensor within the six time periods recorded and analyzed, as shown in Table 1.

Table 1 discharge record table

Using this method to verify that the statistic C ≥ the critical value C1-α(n) within the second time period proves that there is interference. After further on-site analysis, it is judged that the interference comes from the discharge of overhead line clips near the gas-insulated combined electrical appliances in rainy weather. It is precisely because of the existence of this large-amplitude interference signal that the discharge amplitude received by the sensor closer to the discharge point is greater than that of the sensor signal farther away (the external interference and the internal discharge signal are superimposed).

In addition, by comparing the signal fronts of the two sensor waveforms obtained on site, it can be found that the interference signal is later than the arrival time of the internal discharge. This is also consistent with the actual site. Therefore, by screening the abnormal values in the partial discharge early warning data and comparing and analyzing the external interference that occurs during the monitoring process, the location of the interference and its interference frequency band can be found through time domain and frequency domain analysis. Finally, the state of the monitoring device is determined . When the technicians in the monitoring and early warning center find that there is indeed a suspected discharge in the gas-insulated combined electrical equipment, they will immediately issue an early warning notice to the equipment power supply bureau; and cooperate with timely investigation of equipment abnormalities.

The above descriptions are only preferred embodiments of the invention, and are not intended to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present invention shall be included in the protection of the present invention. within range.

Claims (3)

1. A method for screening abnormal values of state early warning data of gas-insulated combined electrical appliances, characterized in that: comprising the following steps: Step 1: Collect the monitoring data of each gas-insulated combined electrical appliance through the partial discharge monitoring devices distributed in various places; Step 2: Upload the monitoring data to the master station system platform data server of the monitoring and early warning center for data analysis; the steps of the data analysis are as follows: (1) Analyze and calculate the monitoring data samples in the data server of the main station system platform, and extract the discharge amplitude monitoring data samples of a set of monitoring devices within a specified time period as X1, X2, X3, ..., Xn; (2) Arrange the monitoring data samples Xi into a statistical data set X(i) in order from small to large, and arrange them randomly if there are monitoring data with the same measured value, that is: X(1)≤X(2)≤...≤ X(n); (3) Solve the statistic C for different sample sizes n, combine the Dixon analysis method to test the consistency of the historical samples of the discharge amplitude in engineering applications, and solve the statistical value of the confidence interval to obtain a significant difference level α The empirical value is 0.01 , to solve its critical value C1-α(n), so that the sample data value can warn the operating state of the gas-insulated combination device; (4) Analyzing and diagnosing the status of the monitored gas-insulated combined electrical appliances based on the screened monitoring data samples; Step 3: Screen and analyze abnormal values in various partial discharge early warning data that characterize the state characteristics of gas-insulated composite electrical appliances through consistency inspection, including the following steps: (31) Extract the discharge amplitude monitoring data samples of a certain partial discharge monitoring device within a specified time period as X1, X2, X3, ..., Xn; (32) Arrange the monitoring data samples Xi from small to large into a statistical number sequence X(i), among them, if there are monitoring data with the same measured value, they will be randomly arranged, that is: X(1)≤X(2)≤...≤ X(n); (33) Solve the statistic C for different sample sizes n, combine the Dixon analysis method to check the consistency of the discharge amplitude historical samples in engineering applications, and set the arithmetic mean value of the discharge amplitude monitoring data sample sequence as but: Therefore, the error of a certain discharge amplitude X _i under the same operating conditions is ΔX, then: If ΔX is too large, the autocorrelation of the error term may be very serious, which will lead to the inconsistency of the estimated parameters, because the discharge amplitude data series is a non-stationary system; test the level of significant difference, and judge whether the confidence is Satisfy; (34) According to the statistical value of the confidence interval, the empirical value of the significant difference level α is 0.01. When the sample interval of the discharge amplitude monitoring data is [C ₁ , C ₂ ], the confidence of the interval [C ₁ , C ₂ ] Confidence interval with a level of 1-α, C ₁ is the measured value of the upper confidence limit, and C ₂ is the measured value of the lower confidence limit; (35) Correspondingly, when the arithmetic mean value of the discharge amplitude monitoring data sample sequence is Solve the significant difference level α empirical value α(n) of the nth sample statistic C, and then solve its critical value In this way, the operating status of the gas-insulated combined electrical appliance can be analyzed through whether the sample data is normal; when the sample statistic C≥ critical value Then the sample x(n) is an outlier; when the sample statistic C<critical value Then the sample x(n) is a normal value; (36) Analyzing and diagnosing the state of the monitored gas-insulated combined electrical appliances and diagnosing the type and severity of partial discharges based on the screened monitoring data samples; Step 4: Accurately identify the typical abnormal value representation in the discharge amplitude, avoid the interference and misjudgment of abnormal values on normal monitoring and early warning, fault diagnosis, and finally reliably evaluate the status of the equipment. 2. A screening method for abnormal values of state early warning data of gas-insulated combined electrical appliances according to claim 1, characterized in that: said partial discharge monitoring device passes the monitoring data collected in gas-insulated combined electrical appliances through wired power integrated data The network is sent step by step to the data server of the main station system platform for screening and analysis. 3. A method for screening abnormal values of state early warning data of gas-insulated combined electrical appliances according to claim 1, wherein the partial discharge monitoring device includes a UHF signal sensor, a signal collector and a comprehensive processing unit, the The monitoring data output by the UHF signal sensor is monitored and analyzed sequentially through the signal collector and the comprehensive processing unit.